https://www.irjet.net/archives/V6/i11/IRJET-V6I11183.pdf

1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 06 Issue: 11 | Nov 2019 www.irjet.net p-ISSN: 2395-0072
© 2019, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 1794
Analysis on Existing Methodologies of User Service Rating
Prediction Systems
Dr. R. Gunasundari1, Sneha Prakash2
1Associate Professor, Dept. of Computer Science and IT, Karpagam Academy of Higher Education, Echanari,
Coimbatore, Tamil Nadu, India
2Research Scholar, Dept. of Computer Science and IT, Karpagam Academy of Higher Education, Echanari,
Coimbatore, Tamil Nadu, India
------------------------------------------------------------------------***-------------------------------------------------------------------------
ABSTRACT:- Recent advancements in the field of mobile and
social networking lead to the increase in usersopinionsharing
by providing reviews, ratings, photos, check INSetc. Theseuser
data can be analyzed to get ideas about their personal
behavior there by helping to predict their required future
service patterns. The use of mobile phones with network
connectivity helps us to geographically locate the smart
phones thus bridges the gap between physical and digital
world. These location based data actsasaconnection between
user’s physical and virtual behaviors. These social networks
involving geographical information are called as Location
based social networks (LBSNs)
Key Words: service rating Prediction, user rating,
Recommendation systems, and social networks
1. INTRODUCTION
Now a days due to wide spread availability of internet
access and use of different mobile devices,social media such
as face book , twitter , LinkedIn are used commonly .These
smart phone users produce large volumes of data . The
internet revolution has brought about a new way of
expressing an individual's opinion. It has become a medium
through which people openly express their viewsonvarious
subjects. These opinions contain useful information which
can be utilized in many sectors which require constant
customer feedback. The proposed method compares about
existing methodologies available for user service rating
prediction and their problems.
2. RECOMMENDER SYSTEMS
Recommender systems are the main methods that helps
users to make decision about what products to buy, which
hotels are good, which movies to watch etc.These systems
helps the business by increasing sales and also helps the
users by saving time and money by recommending the best
personalized results.
Three categories of recommender systems have been
identified which are collaborative filtering, content based
Filtering and the hybrid version which is the combination of
two mentioned categories. Social recommendersystems are
actually the improved version of collaborative filtering.
These collaborative filtering is based on social networks.
Social networks are made of a finite group of users
And their relationships (figure1) that they establish among
Them through the social links so one key insight is that
Social-based recommender systems should account for a
Number of dimensions within a user’s social network,
Including social relationship strength, expertise, and user
similarity.[1].
Figure 1: the pattern of social relationship in social
Networks [2].
3. COLLABORATIVE FILTERING
Collaborative filtering is the technique used by
Recommender systems to filter the user’s behavioral
patterns and extract new patterns.Thisextractionisdone by
collecting users Preferences or interest from many users
.The collaborative filtering method works under the
assumption that if a user 1 has the same opinion of user 2 in
a issue then they are more likely to have same opinion on a
different issue .so by analyzing the opinion of user 1 we can
predict the opinion of user 2. The main pillars of this
approach are that a) many peoples must participate) there
must be a method for people to express their interest c) an
algorithm must be there to match the user’s interest.
Collaborative filtering has found its applications on theweb.
Electronic commerce sites such as Amazon.comandFlipkart
feature recommendation centers, where, in addition to
expert reviews, users can rate items and then receive
personalized recommendationscomputedbya collaborative
filtering engine. User preference can also infer from site
usage: for example, purchasing a book may be taken as
evidence of interest not just in that book, but also in the
book’s author [3].

2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 06 Issue: 11 | Nov 2019 www.irjet.net p-ISSN: 2395-0072
© 2019, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 1795
There exist Different collaborative filtering systems in both
industry and academia. Tapestry [4], a manual collaborative
filtering system presented by Goldberg et al.Relied on each
user to identify like-minded users manually. It wasdesigned
to recommend documents drawn from newsgroups to a
group of users. GroupLens [5,6], Video Recommender [7],
and Ringo [8] were thefirstautomatedcollaborativefiltering
system.
According to Breese et al [9], algorithms for collaborative
recommendations can be grouped into two general classes:
model-based and memory-based Memory based algorithms
are heuristics that use previously rated items by theusers to
make rating predictions. In contrast to memory based
methods, model-based algorithms first uses a set of ratings
to learn a model, then using this model make rating
predictions.
A. Problems
Collaborative filtering suffers from problems such as
Cold start, Scalability, Sparsity [10] [11].
 Cold Start: For a new user or item, there isn't enough
data to make accurate recommendations.
 Scalability: In areas where these systems make
recommendations, there are millions of users and
products. Thus, a large amount of computation poweris
often necessary to calculate recommendations.
 Sparsity: The number of items sold on major e-
commerce sites is extremely large.Themostactiveusers
will only have rated a small subset of the overall
database. Thus, even the most popular items have very
few ratings.
4. CONTENT-BASED RECOMMENDATION
Content based Recommendation is based on Description of
the products and a Profile of user’s interest and preferences.
These methods are suitable to apply when details about an
item like name, location, description, rate etc are known. In
content based Recommendation Systems the user specific
recommendations are made based on user’s likes and
dislikes. These methods try to recommend items that are
similar to items that users liked in the past.
This method sets up a user Profile and item profile. User
Profile for every user based on information's of users like a
pattern of user’s preference and history of users which we
derive from the interaction of users with recommender
Systems. Item profile contains the features, ratings and
details about the item.
Content based system also makes useofuseropinionswhich
will be got from analyzing the feedback collected from users
by leaving a space for users to enter feedback while
purchasing the item. So in turn this system makes use of
opinion mining method.
The existing Content based recommendation systems are G.
Salton, in book “Automatic Text Processing” [12] has shown
that the content-based recommendation has its roots in
information retrieval and information filtering. It is due to
the early advancements made by the information retrieval
and filtering communities that many content-based
recommenders focus on recommending items containing
textual information such as URLs and documents.
NewsDude [13] presented by Billsus & Pazzani (1999), a
content-based filtering system suggests newstoriestheuser
might like to read. To accomplish this two user models are
built. The first user model measure similarity between the
new story and the stories that the user has read before by
counting the co occurrences of words appearing in these
stories. The second user model assigns a probability of
interest to a new story by comparing how frequently its
words occur in those stories the user regards as interesting
to those the user regards as of no interest. Comparing how
frequently its words occur in those stories the user regards
as interesting to those the user regards as of no interest.
5. Hybrid recommendation
One way to implement Hybrid recommender systems are
implemented by combining the results of collaborative
filtering and content based recommender systems. Initially
the data is processed separately by the two methods and
then the results are combined. The other methods include
There is no reason why several different techniques of the
same type could not be hybridized. Hybrid approaches can
be implemented in several ways: by making content-based
and collaborative-based predictions separately and then
combining them; by adding content-based capabilities to a
collaborative-based approach or by unifyingtheapproaches
into one model. Several studies that empiricallycomparethe
performance of the hybrid with the pure collaborative and
content-based methods and demonstrated that the hybrid
methods can provide more accurate recommendations than
pure approaches. These methods can also be used to
overcome some of the common problems in recommender
systems such as cold start and the sparsity problem, as well
as the knowledge engineering bottleneck in Knowledge
based approaches.
The hybrid recommender systemsalsohavesomeissueslike
they need a database of user ratings as well as user profiles.
Without this database recommendersystemscannotpredict
accurate results
6. CONCLUSION
In our opinion, the rating behavior in recommender system
could be analyzed in these aspects: 1) when user rated the
item, what the rating is, 2) what the item is, 3) what the user
interest that we could dig from his/her rating recordsis,and
4) how the user’s rating behavior diffuses among his/her
social friends.

3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 06 Issue: 11 | Nov 2019 www.irjet.net p-ISSN: 2395-0072
© 2019, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 1796
7. REFERENCES
1. Maryam Nayebzadeh Akbar Moazzam, Amir
Mohammad Saba, Hadi Abdolrahimpour, Elham
Shahab, "An Investigation on Social Network
Recommender Systems and Collaborative Filtering
Techniques".
https://www.researchgate.net/publication/318849
587_An_Investigation_on_Social_Network_Recomme
nder_Systems_and_Collaborative_Filtering_Techniq
ues
2. W.-K. Chen, Al Falahi, Kanna,NikolaosMavridis,and
Yacine
3. Atif."Social networks and recommender systems: a
world of current and future synergies." In
Computational Social Networks, pp. 445-465.
4. Springer London, 2012
5. Beyond Recommender Systems: Helping People
Help Each OtherB. Smith, Loren Terveen and Will
Hill AT&T Labs – Research in HCI in the New
Millennium, Jack Carroll, ed., Addison-Wesley,2001
p. 1 of 21
6. Hofmann, T. Collaborative Filtering via Gaussian
Probabilistic Latent Semantic Analysis. In Proc. of
the26th Annual International ACM SIGIR.
Conference, pp.259-266, ACM press, New York, NY,
USA, 2003.
7. Konstan, J. A., B. N. Miller, D. Maltz, J. L. Herlocker,
L.R. Gordon, and J. Riedl. GroupLens: Applying
collaborative filtering to Usenet news.
Communications of the ACM, vol. 40, no. 3, pp. 77-
87, 1997
8. Resnick, P., N. Iakovou, M. Sushak, P. Bergstrom,
andJ. Riedl. GroupLens: An open architecture for
collaborative filtering of netnews. In Proceedingsof
the ACM conference on Computer Supported
Cooperative Work, pp. 175- 186, New York, NY,
USA, 1994.
9. Hill, W., L. Stead, M. Rosenstein, and G. Furnas.
Recommending and evaluating choices in a virtual
community of use. In Proceedings of CHI’95 SIGCHI
Conference on Human Factors in Computing
Systems, pp. 194-201, ACM press, New York, NY,
USA, 1995.
10. Shardanand and P. Maes. Social information
filtering: Algorithms for automating ‘word of
mouth’. In Proceedings of SIGCHI Conference on
Human Factors in Computing Systems,pp.210-217,
ACM press, NY,USA, 1995.
11. Breese, J. S., D. Heckerman, and C. Kadie. Empirical
analysis of predictive algorithms for collaborative
filtering. In Proceedings of the Fourteenth
Conference on Uncertainty in Artificial
Intelligence,pp. 43- 52, Morgan Kaufmann
Publishers, San Francisco, CA, USA, 1998
12. Rubens, Neil; Elahi, Mehdi; Sugiyama, Masashi;
Kaplan, Dain (2016). "Active Learning in
RecommenderSystems". InRicci, Francesco;Rokach,
Lior; Shapira, Bracha (eds.). Recommender Systems
Handbook (2 ed.). Springer US. doi:10.1007/978-1-
4899-7637-6_24. ISBN 978-1-4899-7637-6
13. Elahi, Mehdi; Ricci, Francesco; Rubens, Neil
(2016). "A survey of active learning in collaborative
filtering recommender systems". Computer Science
Review. 20: 29–
50. doi:10.1016/j.cosrev.2016.05.002
14. Salton, G.Automatic Text Processing. Addison-
Wesley Longman Publishing co., Boston, MA, USA,
1989.
15. Billsus, D. and M. Pazzani. User modeling for
adaptive news access. User Modeling and User-
Adapted Interaction, vol. 10, no. 2-3, pp. 147-
180,Kluwer Academic Publishers Hingham, MA,
USA,2000.